ABSTRACT
With the rise of COVID-19, decision support systems (DSS) increasingly display crowding information (CI) (e.g. how crowded a medical practice is) to encourage physical distancing when users select locations. Despite important implications for containing COVID-19, little is known about the causal effect of CI on user selection behaviour and how the immediacy of CI (e.g. “updated 2 minutes ago“) as well as users’ health anxiety further influence the effect of CI. Drawing on literature on digital choice environments and construal level theory, we conducted a multi-national online experiment to investigate the effect of CI on selecting differently crowded medical practices. Our results demonstrate that present (vs. absent) CI in DSS increases the likelihood of users selecting less crowded medical practices, while the effect is strongest when employed with real-time (vs. historical average) CI and, surprisingly, when users’ health anxiety is low (vs. high). Overall, our study adds to the growing body of research on IS in the age of pandemics and provides actionable insights for DSS providers and policymakers to endow users with information to identify and select less crowded locations, thus containing COVID-19 through improved physical distancing without paternalistically restricting users’ freedom of choice.
SPECIAL ISSUE EDITORS:
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
2. Often referred to as “social distancing“ in media. We refrain from using the term social distancing, because information and communication technology-mediated social interactions (e.g. phone, chat, video streaming) are not restrained by the COVID-19 pandemic.
3. As of June 20th, 2020, Italy reported 58 COVID-19-related deaths per 100,000 citizens, whereas Germany reported only 11 COVID-19-related deaths per 100,000 citizens (Johns Hopkins University, Citation2020).
4. In a separate analysis, we find that processing fluency is significantly higher for real-time rather than historical average CI (p < 0.001), indicating support for our proposition of enhanced construal fit.